ARM(ARM.US) 3QFY25 performance meeting: royalty revenue increased by 23% year-on-year, expected to become a driving force for future income growth.

On the morning of February 6, 2025, after the US stock market closed, ARM (ARM.US) released its financial report for the third quarter of the 2025 fiscal year (ending in December 2024) and held a conference call to explain the third quarter performance. Company executives stated that the total revenue for the third quarter was $983 million, exceeding the high end of expectations, with a year-on-year growth of 19%. Breaking it down, royalty revenue reached a record $580 million, a 23% increase year-on-year, exceeding expectations. The growth was mainly attributed to the continued adoption of the Armv9 architecture and chip shipments based on the Compute Systems Subsystem (CSS). The company stated that both v9 and CSS do not lead to a one-time increase in royalty rates, but rather a sequential change, with an annual increase typically accompanying the launch of new versions. It is expected that most of the future growth will come from royalty revenue rather than licensing revenue. ARM stated that royalty revenue for the quarter increased by 23%, and the total percentage remains steady. However, if only looking at the dollar revenue from v9, the proportion increased from 15% last year to 25% this quarter. Therefore, the absolute dollar revenue increased by three digits. Hence, relative stability is actually a good sign, indicating there is still a long growth runway. The company still expects v9 to grow to 67% to 70% of total royalty revenue. Furthermore, the fact that the company is able to meet or exceed the royalty growth rate while v9 growth is not accelerating should give the company more confidence in its ability to drive further royalty growth in the coming quarters and beyond. Regarding licensing revenue, the quarter recorded $403 million, a 14% increase year-on-year, higher than expected. Licensing revenue fluctuates significantly due to timing and scale of high-value licensing agreements, and it is recommended to pay attention to Annual Contract Value (ACV), which grew by 9% in the third quarter year-on-year. As for the full-year 2025 fiscal year expectations, company executives stated that the midpoint of the full-year revenue expectation has been raised to about $4 billion, a 24% year-on-year growth, exceeding the long-term target of 20%. It is expected that the year-on-year growth rate for royalty revenue for the full year will be in the high teens, and licensing revenue will grow by about 30%. Non-GAAP operating expenses are expected to be $2.1 billion, a 21% increase year-on-year. Full-year non-GAAP EPS is expected to be between $1.56 and $1.64. Q&A: Q Regarding the growth points for licensing revenue in the next year. This question is also related to some recent news we have seen in the AI field, and also involves Arm's capabilities. Two weeks ago, we saw the announcement of the Stargate project, with a budget exceeding $100 billion. Recently, we have heard about the new collaboration project Cristal Intelligence between SoftBank and OpenAI. We are trying to understand Arm's position in these projects, as it seems that Arm is both a driver and has a first-mover advantage in the AI era. Could you please outline more clearly the opportunities for Arm, including the possibility of selling new products for these new projects, and the potential returns for Arm from a long-term perspective? A The Stargate project is an extremely important infrastructure project, with the US immediately investing $100 billion and planning to invest $500 billion in the future. This is a project in collaboration with OpenAI, Oracle, and SoftBank, with technical partners including us, Microsoft, and NVIDIA. For Arm, we are very excited to be the preferred CPU for such a platform. Combined with the Blackwell CPU and Grace, Arm will be the initial choice for this configuration. Looking ahead, there is tremendous potential for technological innovation in this field. This is a very exciting project, and we believe it will have a transformative impact on the industry. The Cristal Intelligence project focuses on agent AI, where agents will move on every node of the hardware ecosystem. From the smallest devices (such as earbuds) to data centers, this essentially means agents will become the core interface or DRIVE driving AI inside devices. For Arm, this is a major opportunity as AI workloads will run on every terminal I mentioned. Moreover, given that Arm is the most popular computing platform globally, these AI workloads will run through Arm. Through our KleidiAI library, we will enable developers to easily optimize these agents for the Arm platform. The announcements of these two projects are therefore significant for both the industry and our company. In terms of the driving factors for the growth of royalty revenue, licensing revenue this quarter was about $27 million higher than our expectations, a 14% year-on-year growth. The main DRIVES in the past few quarters are still the demand for AI and the continuous adoption of v9 technology required for AI chips, as well as CSS. As Rene mentioned, we have also signed some additional CSS contracts and we are continuing to push forward. For the upcoming fourth quarter, we expect licensing revenue to increase by about 60% year-on-year. From the beginning of the year, we mentioned that we planned to sign some large contracts in the fourth quarter. These contracts are mostly going according to plan and are meeting expectations. These contracts align well with the categories driven by AI and CSS. Q You mentioned that the demand for CSS from partners is stronger than initially expected. How should we think about the momentum in royalty revenue growth as we enter the 2026 fiscal year compared to your previous expectations? A We are not ready to discuss the situation for the 2026 fiscal year at this time. However, from the exit rate of the previous quarter, we are very satisfied with the current momentum. For example, royalty revenue in the previous quarter reached $500 million, which is about 13% higher than our previous record of $514 million. Of course, this is not surprising as it is also in line with our efforts.In this quarter, we obtained substantial revenue from CSS for the first time, mainly led by the Dimensity 9400 chip we mentioned. In addition, Cobalt CSS has also started shipping and deployment, which is translating into royalty revenue. With the increase in CSS deployments in future quarters, we expect the momentum to continue. As for quantifying this momentum, we will need to wait until next quarter.Q Last month, we saw news reports about a lawsuit between your company and a major client. According to the media reports, there seems to be some disagreement on the expiration date of the contract. They believe the contract will expire in 2028, while we believe it is 2025. I know these are all comments based on the Gregorian calendar years, but can you talk about your current views? For example, do you expect any revenue that will not be realized in the next Gregorian year, or can we quantify it? A This question is about the impact of the lawsuit with Qualcomm on revenue. We predicted from our IPO that we would not win this lawsuit and we have always assumed so. The main reason for this lawsuit is to defend our intellectual property, which is very important. But from a financial perspective, we have always assumed that we will continue to collect royalties at the same rate as in the past, and they will continue to pay. Q I did not see the contribution of v9 in royalty income in the shareholders' letter. Maybe I missed it. If you have the numbers from the last quarter, could you tell me? A As far as total royalty income in the last quarter is concerned, the percentage is 25%, consistent with the previous quarter. Q The 25% figure has been the same for several quarters now. Do you expect it to rise again in the coming quarters? I know the answer may be yes, but... can you help us understand why it has been stagnant at 25% in the past few quarters? A First of all, this is a total percentage, so we saw a 23% increase this quarter, so it has remained steady as a total percentage. But if you only look at the dollar income for v9, its proportion has increased from 15% last year to 25% this quarter. Therefore, the absolute dollar income has grown by three digits. So, the fact that it remains relatively stable is actually a good sign, indicating that we still have a long growth runway. We still expect v9 to grow to 67% to 70% of total royalty income. And we should be able to meet or exceed the royalty growth rate in the upcoming quarters and beyond, even if v9 growth does not accelerate. Additional: Perhaps just to give you a sense of how to think about these changes, they are primarily driven by the shift of OEM products and when they are launched. For example, the MediaTek 9400 has been designed into OPPO and vivo phones, which are now in mass production. So, you will see peaks as production increases, and then it will stabilize. But with the release of new versions, two things will happen. The wider adoption of high-end products and the shift from high-end to mid-high-end, mid-range products from v8 to v9. Therefore, the conversion rate we see is completely in line with our expectations and very consistent with the overall growth trajectory we expected. So, we are very satisfied with the current progress. As you said, there is still room for growth, but this is mainly driven by OEM changing chip combinations, not by new licensed companies. Q Regarding royalty growth, can you talk more specifically about the situation in the 2026 fiscal year, the contribution of v9 throughout the year, and the evolution of the product portfolio? A We will provide guidance for the 2026 fiscal year in the next quarter, as we usually do when announcing fourth quarter performance. So, we will provide more details. I just want to say that we have previously stated that we expect royalty income to grow by about 20%. We have not updated this forecast today. We will achieve this goal through the combination of v9 and CSS. So, what you saw in the last quarter was actually a strong growth in v8. This is good, but overall, you should expect v9 to continue to grow. But also remember, v9 is just one factor in growth. CSS is a bigger growth factor, as the royalty rate for CSS is about double that of v9, and of course, the royalty rate for v9 is about double that of v8. So, you need to consider all factors, but based on the momentum of the last quarter and our expectations for the next quarter, we believe this will set a good foundation for the 2026 fiscal year. Additional: The CSS royalty rate for the 2026 fiscal year may not be the same as the CSS royalty rate for the 2025 fiscal year. These CSS rates will change over time and increase as new solutions are introduced to the market. So, these are subject to change over time. Q License revenue is obviously difficult to predict specific weekly, daily, or monthly figures. But are there some structural changes happening, such as the complexity of your customer agreements, which mean these agreements will take longer and longer to complete? Should we expect this trend in the next few years? A You need to categorize contracts mainly into two types. One is brand-new contracts with companies we have never worked with before. These may take longer. The other is renewals of existing ATA (Technology License Agreement), which is usually a shorter period. But in any case, larger deals often take longer. When transaction amounts reach billions of dollars, they may involve different approval processes like boards, which may take many months. But overall, in response to your question, will there be any changes next year or the year after compared to the past year? No, there will be no change. Q Regarding AWS, you have achieved very good success with AWS, with almost 90% of the top 1000 customers using Graviton and Arm IP. Can you talk about the progress of Cobalt 100, how fast it is developing? I have a follow-up question. A I will leave specific comments on Cobalt shipments to Microsoft. But overall, what we see is not only an increase in the deployment momentum of these products in the cloud. And not only has Grace Blackwell, pivoting from x86 to H100 shown higher growth. This is good, but overall, you should expect v9 to continue to grow. But also remember, v9 is only one element of growth. CSS is a bigger growth factor, as the royalty rate for CSS is about double that of v9, and of course, the royalty rate for v9 is about double that of v8. So, you need to consider all factors, but based on the momentum of the last quarter and our expectations for the next quarter, we believe this will set a good foundation for the 2026 fiscal year. Additional: the CSS royalty rate for the 2026 fiscal year may not be the same as the CSS royalty rate for the 2025 fiscal year. These CSS rates will change over time and increase as new solutions are introduced to the market. So, these are subject to change over time.The transformation of the GB200 also provides support for Arm's development in the data center. Because the main CPU in these AI data centers is Grace, from the perspective of host control, this is a good tailwind. Therefore, we are very satisfied with the overall momentum. I will leave specific comments about Cobalt to Microsoft again, but the momentum has been very good.I believe this will be quite stable. The majority of our related parties are Arm China, so we expect this to remain relatively stable. Arm China may account for a smaller percentage of revenue. I think it's around 25% this quarter. In the past, we have said that we expect it to decrease to the mid teens percentage over the next few years. But there may not be significant changes in the next few quarters. Regarding mergers and acquisitions, I cannot comment on any speculations or rumors. Most of my time is spent thinking about the future of CKH HOLDINGS. We are all thinking about the speed of AI development and the changing speed of software models, which puts a lot of pressure on our ecosystem, requiring us to develop products faster and better. As Arm, we are at the core of all this, striving to solve our customers' problems. Unfortunately, I cannot provide more details or speculate on any acquisition rumors. We have indeed seen an acceleration in the demand for CSS. Customers clearly see its advantages. We are now involved in CSS in almost all major markets we participate in. We have also launched the Arm Total Design Partner Program. This allows end partners, design companies to adopt our CSS, and potentially develop chiplets, with a strong demand for this project. The complexity of these designs is continuously increasing, and the relationship between hardware and software is very close. As Arm is at the center of the software ecosystem in these chips, the demand for us is significant, requiring us to do more things faster to help products enter the market faster. We are in a great period in the industry where computing demand is exceeding the silicon that serves it. We receive many questions about the smartphone market and AI capabilities to leverage what is happening. You have to remember that the chips in these smartphones were designed 2 to 3 years ago, and the memory subsystems, power, etc., were all predefined. Therefore, adapting these small language models or anything to smartphones, given that you still need to run displays, operating systems, and applications, is quite a challenge. So helping solve this problem and getting products to market faster is something we are very focused on. So, what you really want to ask is if this sets the stage for growth next year? I think it can. I just don't know if it will be next year or the year after, or exactly when. As you said, our previous assumption was that it would grow from 10% to 15%, then to 20%, and then to 25%. With several quarters stabilizing at 25%, we provide this indicator to help people understand the adoption of v9. The slowdown in the growth of v9 as a percentage of total royalties is actually a good thing, as it provides us with better prospects for growth because we still believe it will reach 60% to 70%. Why do we believe this? Well, we have signed contracts, and we already know the v9 products that are going to market. So, ultimately, these are the driving factors. Again, we provide this number to help people understand the leading indicator of royalty growth. So, we are able to meet or exceed royalty figures even as v9 adoption slows down, which is actually a good thing. But I think you should focus more on overall royalty growth rather than v9 adoption. When we provide this indicator, the rate of royalty growth is much slower now. So, I think you should focus more on overall royalty growth rather than v9 adoption. This is a more macro question. Recently, there have been many investors discussing issues related to training optimizations, as well as discussions about DeepSeek. Assuming these are effective, does it change your view on the opportunities for Arm in the field of AI? More broadly, I would like to hear your thoughts on this topic, not just about Arm.Whether it is their V3 general model or R1 reasoning model, it is obvious that they have done a lot of work based on existing work (i.e., cutting-edge models created by others), and then DeepSeek has done a very creative job, establishing a highly efficient reasoning model.I think this is good for the industry because it improves efficiency and reduces costs. By doing so, it increases the overall demand for computing. So, from a general perspective, this is a good thing. When you consider the applications of Arm, given that AI workloads need to run anywhere, more efficient inference reduces costs, making it easier to run these applications in areas where power and computing capabilities are limited. For Arm, this is very good. Products like Grace Blackwell, while great, can never be put into a phone, a headset, or even a car. But Arm is everywhere. Therefore, I think it is very good for Arm when reducing the overall inference costs. I also think we are still far from being good enough, because if you look at the spending announced by the big players in this market, Alphabet Inc. Class C just announced they will spend $60 billion in their earnings call. Satya (Microsoft Corporation CEO) mentioned $80 billion. Meta talked about $40 to $50 billion. No one is backing down. The reason is that we are still far from the transformative capabilities that AI can achieve. So, I actually think this is a good thing because it will increase the overall demand for computing. For Arm, this is even better as it allows us to play a role in areas where efficiency is crucial, which is our strength. Q Following up on the question of Annual Contract Value (ACV). You mentioned a slowdown in growth. I know you will provide full-year guidance later. Just curious, how do you see ACV in the long term? Are there any large data center or mobile projects coming up that could lead to double-digit or high single-digit growth? How much of ACV comes from Arm China business? A Regarding our forecasts and where we believe it should be, if you go back to when we went public, we basically provided ACV and thought it should grow between mid to high single digits. The reason being, if you consider all our deals as allocable, most of our licensing revenue and ACV actually come from ATA deals. And almost all ATA deals usually have a 7% annual growth. So it should be... and all non-ATA deals won't have that. So it might be around 7% up or down. So that's our forecast. And over the past, now it's been 6 quarters, due to AI, we have seen an acceleration of licensing revenue, mainly around the adoption of v9 and now CSS. So this provides us with upside ranging from that 7% mark up to 14% or 15% high point. Will we get back to the mid-teens range? I don't know. That's really something we've never really forecasted. When I provide guidance next quarter, I will try to give you more insights. But in the long-term growth model, you should expect most of the growth to really come from royalties. So, of course, the royalty rates for v9 are much higher. Both v9 and CSS are not one-time increase in royalty rates. They are step-change increases, but there's usually an annual increase with the release of new versions. So in the future, you should expect that most of the growth will come more from royalties, not licensing revenue. Q How much of ACV comes from Arm China? A The Arm China portion is about 20%, so it's fairly close to the overall mix. Last quarter it was 25%, so it's in that range. Q A question about CSS, is the CSS licensing activity primarily in data centers or is it balanced between data centers and mobile? And Rene, I just want to confirm what you mentioned before. When AI enters smartphones, does this mean that these processors will adopt chiplet architecture? So in the long term, could the attach rate of CSS in smartphones be very high? A I think in every endpoint you described, earbuds may not be suitable because they are too small. But this chiplet approach will be prevalent in almost every SoC. In other words, within the package, you will see multiple small chips everywhere. It's already almost a standard in high-end. But I think you will see it everywhere, and this is a huge opportunity for us because it not only provides computing CSS from a CPU perspective, but also allows us to have the right combination, be it with NPU or GPU, and the right CPU combination, to maximize performance. You actually touched upon a very crucial demand-driving factor, which is the reason we see strong demand for CSS types in all these endpoint markets. Q And in the Internet of Things, will earbuds also become CSS devices? Or is it a single-chip wafer with just more processing power on an Arm chip? A Regarding the combination of CSS, we have talked in the past that in the automotive space we have already announced but it's not here yet. So from the roughly dozen CSS licenses we have sold in the past, assuming it's basically 50-50 between infrastructure and customer businesses. Additionally: I want to emphasize that this chiplet architecture will be everywhere. This is a huge opportunity for us as we can provide computing CSS from a CPU perspective and combine NPU or GPU as needed, as well as the right CPU combination for maximizing performance. This architecture will be widespread in almost every SoC, which is why we see strong demand for CSS in all endpoint markets.